Assessing Depth of Small Bowel Insertion at Push Enteroscopy by Using Capsule Endoscopy

Consecutive adult patients (aged > 18 years) referred to a single tertiary referral center for investigation of OGIB will be prospectively recruited. Patients will be included if they had overt signs of GI bleeding (hematochezia or melena with an associated drop in hemoglobin) or occult signs of GI bleeding in the form of persistent or recurrent iron deficiency anemia and had undergone esophagogastroduodenoscopy (EGD) and colonoscopy that were negative for a source of bleeding. Exclusion criteria include significant cardiopulmonary disease, an implanted electromedical device, pregnancy, previous gastrectomy, extensive Crohn's disease with suspicion for stricturing or fistulizing small bowel disease. All patients will provide informed consent before enrollment.

The study protocol was approved by the Institutional Review Board (IRB) at our institution and was conducted in accordance with ethical guidelines of the Helsinki Declaration Procedures Patients will undergo bowel preparation by oral ingestion of 2L of polyethylene glycol solution 12 hours before the study. Push enteroscopy (PE) will be performed in an endoscopy theater with patients under conscious sedation using intravenous injections of midazolam (0.05-0.1mg/kg) and fentanyl (1-2 mcg/kg), with dimenhydrinate (25-50mg/body) added in the case of insufficient sedation. Cardiorespiratory function will be monitored during the procedure. Antiplatelet and anticoagulant medications will be discontinued 3-7 days before the procedure. No intravenous antispasmodic will be used during the PE in efforts to preserve normal gut motility for the passage of the subsequent capsule endoscopy (CE). A pediatric colonoscope furnished with a water-jet system (PCF-TYPE H190L; Olympus Co, Tokyo, Japan) and distal transparent attachment cap will be inserted through the mouth and be advanced as far as possible into the small bowel using a water immersion technique and aided by multiple reduction maneuvers. No overtube will be used and carbon dioxide will be used for insufflation in all cases. Two endoscopic clips (SureClip; Micro-Tech Co, Nanjin, China) will be placed at the maximal distal insertion point to mark this location for identification by CE. The endoscope will be then withdrawn gradually as per our standard practice for direct examination of the small bowel traversed by the PE. The presence of any vascular or non-vascular lesions will be recorded and endoscopic treatments such as argon plasma coagulation (APC), endoscopic clips or biopsies will be performed as clinically appropriate depending on the nature of the lesions found. In particular, the number of vascular lesions seen and how many of these lesions were treated with APC will be recorded. After PE is completed, the capsule (PillCam TM SB3 capsule; Medtronic, Dublin, Ireland) synchronized with the recording device will be deployed into the first part of the duodenum by using a gastroscope (GIF-HQ190; Olympus, Tokyo, Japan) and capsule deployment device (AdvanCE Endoscope Delivery Device; STERIS, OH, USA). Capsule recording devices will be retrieved 8-12 hours after capsule deployment and images read by two separate endoscopists trained in reporting CE. CE recordings will be reviewed using the RAPID 8.0 software. The location of the clip placed at the deepest insertion point of the PE as seen by CE will be quantified by the percentage of small bowel progress (SBP) and small bowel transit time (SBTT) from the duodenal bulb to the clip relative to the entire length of the small intestine from the duodenal bulb to the cecum. Any bleeding lesions and their respective locations will also be recorded. Furthermore, the number of angioectasias treated during PE that were seen again on CE will be recorded to calculate the detection rates of these lesions.

The primary outcome of the study is to estimate the proportion of the small intestine that can be evaluated by using the PE technique. Secondary outcomes include the diagnostic yield of CE and PE in identifying the cause of obscure gastrointestinal bleeding, comparison of the detection rates between the two modalities and adverse events related to these procedures. Statistical analyses will be performed using JMP Pro 15 (SAS Institute, Cary, NC, USA). Continuous variables will be expressed as a mean with standard deviation (SD) or 95% confidence interval (CI), or median with interquartile range (IQR) according to distribution.